Immunogenicity of Hbv Vaccine During Stated Shelf-Life

SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

IMMUNOGENICITY OF HBV VACCINE DURING STATED SHELF-LIFE

Nina G Gloriani1, Karthik Srinivasa2, Hans L Bock2 and Bernard Hoet2

1Department of Microbiology, College of Public Health, University of the Philippines, Ermita, Manila, Philippines; 2GlaxoSmithKline Biologicals, Wavre, Belgium

Abstract. Thiomersal has been used as preservative in multi-dose vials of hepatitis B vaccine (Engerix™-B). Due to safety concerns, thiomersal was replaced with 2-phenoxyethanol (2PE) as preservative in multi-dose vials. The potency of 2PE preserved hepatitis B vaccine multiple use vials was measured over the shelf-life in terms of immunogenicity, reactogenicity and safety. This single-blind, randomized study was conducted with the assistance of employees of GlaxoSmithKline Biologicals, makers of the Engerix™-B vaccine. Four hundred twenty subjects aged ≥18 years were randomized to receive three doses (0, 1, 6 months) of 2PE preserved hepatitis B vaccine kept on the shelf <12 months (2PE New group), 2PE preserved hepatitis B vaccine kept on the shelf >18 months (2PE Old group), or thiomersal preserved hepatitis B vaccine [HBV(Thio) group]. Anti-HBs was measured by GlaxoSmithKline Biologicals post-vaccination; the reactogenicity and safety of the vaccines were assessed. Protective anti-HBs levels ( ≥10 mIU/ml) were measured one month after dose 3. The results showed protective levels in 86.8% (2PE New), 89% (2PE Old) and 95.3% [HBV(Thio)]. There was no difference de- tected between the 2PE New and 2PE Old groups in terms of anti-HBs seroprotection rates and geometric mean concentrations one month after dose 3. However, both 2PE groups had significantly lower seroprotection rates than the HBV(Thio) group and the number of non-responders was higher in the 2PE groups than in the Thio group. A antibody response rates over time were similar between the 2PE New and Old groups. The reactogenicity profiles were acceptable and the ranges were similar for each group. The shelf-life of the vaccines had no impact on immunogenicity or reactogenicity and 2PE preserved hepatitis B vaccine can be considered stable over time. Key words: HBV vaccine, immunogenicity, shelf-life, preservative

INTRODUCTION and chronic liver disease (WHO, 2004). According to World Health Organization Hepatitis B is a worldwide health (WHO) estimates, two billion people glo- problem and an important cause of acute bally have serological evidence of past hepatitis B infection and approximately Correspondence: Nina G Gloriani, College of Public Health, University of the Philippines 360 million are estimated to be chronically Manila, 625 Pedro Gil Street, Ermita, Manila, infected with the virus (WHO, 2000, 2004). Philippines 1000, Philippines. Of these, 15-25% are expected to die from Tel: +63 2 524 2703; Fax: +63 2 521 1394 liver disease, cirrhosis or primary hepato- E-mail: [email protected] cellular carcinoma (WHO, 2000, 2004). Pre-

876 Vol 41 No. 4 July 2010 STABLE IMMUNOGENICITY OF HBV VACCINE vention of infection through vaccination another preservative, 2-phenoxyethanol is the most effective strategy to reduce glo- (2PE), also commonly used in human vac- bal morbidity and mortality due to hepa- cines. titis B (Zanetti et al, 2008). When stored over a long period of Engerix™-B, GlaxoSmithKline Biolo- time, vaccines are known to gradually lose gicals’ recombinant DNA hepatitis B vac- their potency and have lower immunoge- cine, contains hepatitis B surface antigen nicity (WHO, 1998). This study evaluated (HBsAg) from genetically engineered the immunogenicity, reactogenicity and yeast (Saccharomyces cerevisiae). Immuno- safety of two formulations of multi-dose genicity, safety and efficacy of Engerix™-B Engerix™-B; one containing thiomersal have been demonstrated (Poovorawan et and the other containing 2PE as preserva- al, 1990, 1992; Keating and Noble, 2003). tive. The 2PE containing vaccine was di- Thiomersal has been used as a preserva- vided into two lots: one lot on the shelf tive in multi-dose vials of Engerix™-B. <12 months and one lot on the shelf >18 Thiomersal is an organomercurial preser- months. We report here immunogenicity vative commonly used in vaccines and of the 2PE-containing vaccine at different other parenteral medications. Multi-dose lengths of time on the shelf. vials of hepatitis B vaccine have lower manufacturing costs, require less storage MATERIALS AND METHODS space when compared to monodose vials, and can be used in settings where cold Study design and subjects storage capacity is limited (EMEA, 2001; This was a single-blinded, random- Clements and Wesselingh, 2005). How- ized (1:1:1) study (eTrack study ID: 103860/ ever, unlike monodose vials that are dis- 283) conducted in the Philippines between carded immediately after single use, multi- May 2003 and March 2004 in accordance dose vials are used more than once after with the Good Clinical Practice and Dec- the vials are opened. Thus, preservatives laration of Helsinki (1996 version). The are added to multi-dose vaccine vials to study protocol was approved by the local prevent microbial contamination between ethics committee and written informed use during the vaccine’s shelf-life. consent was obtained from all study par- Thiomersal was previously the preserva- ticipants. tive of choice. The thiomersal present in Subjects aged ≥18 years were enrolled Engerix™-B was a result of the purifica- and randomized into one of the three µ tion process (residual amount of <2 g/20 groups: the 2PE New group (vaccinated µ g HBsAg) and was also added as a pre- with a 2PE-containing hepatitis B vaccine servative to the final vaccine. on the shelf <12 months), the 2PE Old In response to safety concerns about group (vaccinated with a 2PE-containing thiomersal in human vaccines, regulatory hepatitis B vaccine on the shelf >18 agencies called for removal of thiomersal months), the HBV(Thio) group (vaccinated from these products (AAP, 1999; EMEA, with thiomersal-containing hepatitis B 1999). All Engerix™-B vaccines are now vaccine). Subjects received three vaccine manufactured in monodose vials and are doses at 0, 1 and 6 months. Enrolled sub- preservative free (CDC, 2000). However, at jects were seronegative for HBsAg, anti- a time when multi-dose preparations were hepatitis B core antigen (anti-HBc) anti- still in use, thiomersal was replaced with bodies and anti-hepatitis B surface antigen

Vol 41 No. 4 July 2010 877 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

(anti-HBs) antibodies. Subjects were ex- safety cohort, which included all subjects cluded from the study if they were preg- who had not received a vaccine forbidden nant, immunocompromised, had a history by the protocol. The analysis of unsolic- of hepatitis B or had previously received ited symptoms and SAEs was conducted hepatitis B vaccination. on the total vaccinated cohort. Vaccines Subjects with an anti-HBs antibody The vaccine was in a multi-dose (10 concentration ≥3.3 mIU/ml were considered doses) vial. Each 1 ml dose contained to be seropositive and those with a concen- 20 µg of HBsAg. The 2PE formulations tration ≥10 mIU/ml were considered each contained 5 mg of 2PE and <2 µg of seroprotected. Anti-HBs seroprotection thiomersal, while the HBV(Thio) formu- rates and geometric mean concentrations lation contained 50 µg of thiomersal. (GMCs) were tabulated with 95% confidence intervals (CI). Non-inferiority be- Assessment of immunogenicity, reacto- tween the 2PE New group and the genicity and safety HBV(Thio) group (primary study objective) Blood samples were collected at was concluded if the lower limit of the Months 1, 2, 6 and 7 to measure anti-HBs asymptotic 95% CI on the group difference antibodies. The assessment was done at [2PE New minus HBV(Thio)] in terms of GSK Biologicals central laboratory (en- seroprotection rates was >10%. Similar zyme immuno assay, AUSAB® EIA/Abbott non-inferiority calculations in terms of laboratories). seroprotection rates were performed for the Solicited local (pain, redness and group difference between 2PE Old and swelling at the injection site) and general HBV(Thio) groups and between 2PE Old (fatigue, fever, gastrointestinal symptoms and 2PE New groups. Exploratory analy- and headache) symptoms were recorded ses assessed the non-inferiority of the 2PE for 4 days (Day 0 to Day 3) after each vac- New to the HBV(Thio) group, non-inferi- cine dose. Symptom intensity was graded ority of the 2PE Old to the HBV(Thio) group on a scale of 0-3. Grade 3 redness and and non-inferiority of the 2PE Old to the swelling was defined as injection site di- 2PE New group in terms of anti-HBs anti- ≥ ameter 50 mm and grade 3 fever as an body concentrations obtained at one month axillary temperature >39ºC. For all other after dose 3 by computing the 95% CI for symptoms, grade 3 was defined as a symp- the pair-wise GMC ratios among the three tom that prevented normal everyday ac- groups using a one-way ANOVA model. If tivities. Unsolicited symptoms were re- the 95% CI of the GMC ratios between the corded for 31 days (Day 0 to Day 30) after groups under consideration contained the each vaccine dose and serious adverse value 1, then non-inferiority was estab- events (SAE) were recorded throughout lished. the study period. Statistics RESULTS The immunogenicity analysis was performed according-to-protocol (ATP) Subjects cohort that included all eligible subjects A total of 420 subjects were enrolled who complied with the protocol-defined and randomized: 140 in the 2PE New procedures. The analysis of solicited group, 141 in the 2PE Old group and 139 symptoms was conducted on the ATP in the HBV(Thio) group. Twenty-five sub-

878 Vol 41 No. 4 July 2010 STABLE IMMUNOGENICITY OF HBV VACCINE

Table 1 Seropositivity, seroprotection rates and GMC (ATP cohort for immunogenicity).

S+ SP Group Timing N GMC (95% CI) n % (95% CI) n % (95% CI)

2PE New One month 129 117 90.7 112 86.8 2,163.7 post-Dose 3 (84.3-95.1) (79.7-92.1) (1,388.9-3,370.6) 2PE Old One month 127 115 90.6 113 89.0 2,621.0 post-Dose 3 (84.1-95.0) (82.2-93.8) (1,701.9-4,036.6) HBV(Thio) One month 128 123 96.1 122 95.3 2,108.3 post-Dose 3 (91.1-98.7) (90.1-98.3) (1,425.3-3,118.7) N, number of subjects with available results S+, seropositivity for anti-HBs antibodies (≥ 3.3 mIU/ml) SP, seroprotection for anti-HBs antibodies (≥ 10 mIU/ml) n, number of subjects with concentration within the specified range %, percentage of subjects with concentration within the specified range 95% CI, 95% confidence interval GMC, geometric mean concentrations in mIU/ml jects withdrew during the study. Two with- rate one month after dose 3 was 86.8% drew due to non-serious adverse events (95%CI 79.7, 92.1) in the 2PE New group, (prolonged menses and menstrual-like 89.0% (95%CI 82.2, 93.8) in the 2PE Old bleeding) and two withdrew due to SAEs group and 95.3% (95% CI 90.1, 98.3) in the (nervous breakdown, severe headache and HBV(Thio) group. One month after dose numbness of extremities). None of the ad- 3 the antibody GMC was 2,163.7 mIU/ml verse events leading to withdrawal were (95% CI 1,388.9, 3370.6) in the 2PE New considered by the investigator to be related group, 2,621.0 mIU/ml (95% CI 1,701.9, to vaccination. Two subjects were elimi- 4,036.6) in the 2PE Old group and 2,108.3 nated from the ATP safety cohort (n = 418) mIU/ml (95% CI 1,425.3, 3,118.7) in the due to randomization failure and the vac- HBV(Thio) group. There were 12 subjects cine not administered as per the random- each in the 2PE New and 2PE Old groups ized group. Additionally, nine subjects and five in the HBV(Thio) group who did were eliminated from ATP immunogenic- not respond to the three doses of hepatitis ity cohort (n = 409) due to non-compliance B vaccine (Table 1). If the non-responders with blood-sampling intervals. were excluded, the anti-HBs antibody con- Subjects in the three groups were com- centrations in the three groups appeared parable in terms of age, gender and race. similar (Fig 1). The mean age of the subjects was 33.3 Non-inferiority between the 2PE New years (standard deviation 11.5 years); and the HBV(Thio) groups could not be 65.3% were female, 99.5% of subjects were established since the pre-defined non-in- of Southeast Asian origin. feriority criterion was not met (group difference = -8.5% (95%CI -16.0, -1.6)). Non- Immunogenicity inferiority between the 2PE Old and The anti-HBs antibody seroprotection HBV(Thio) was also not established since

Vol 41 No. 4 July 2010 879 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

100 Groups analysis showed no difference be- 2PE New 90 2 PE Old HBV (Thio) tween groups in terms of the post- 80 dose 3 GMC (95% CI on the group

70 ratios included the value 1 for all comparisons; data not shown). The 60 kinetics of the antibody response 50 over time was similar between 40

Percentage of subjects Percentage the 2PE New and 2PE Old groups 30 (Fig 2). 20 Reactogenicity and safety 10 Pain at the injection site (9.1% 0 to 12.9% of doses) and fatigue (3.9% 1 10 100 1,000 10,000 100,000 1,000,000 to 5.6% of doses) were the most fre- Cut-off = 3.30 Anti-HBs Antibody titers (mIU/ml) quently reported local and general Fig 1–Reverse cumulative curve for anti-HBs concen- symptoms, respectively, for all trations after excluding subjects seronegative three groups (Fig 3). Grade 3 local at one month after dose 3 of the hepatitis B and general symptoms were infre- vaccine (according to the protocol immunoge- quently reported (≤ 0.2% of doses nicity cohort). of all groups). During the 31-day 2 PE Old 2 PE New follow-up period, ≤ 2.7% of doses were followed by unsolicited symptoms. Eight doses were fol- 1,000 lowed by at least one unsolicited adverse events that was considered to be causally related to vaccination. All were reactions at the injection site (one in the 2PE new Anti-HBs antibody GMC (mIU/ml) group, two in the 2PE Old group

10 and five in the HBV(Thio) group). 012345678 Five subjects (one each in the 2PE Months Old and New groups and three in Fig 2–Anti-HBs geometric mean antibody concentra- the HBV(Thio) group) reported 6 tions (GMC) in the 2PE New (open circles) and SAEs during the study. None were 2PE Old (squares) groups (according to the determined by the investigator to protocol immunogenicity cohort). Months 1,2 and 7, one month after doses 1,2 and 3 respec- be causally related to vaccination. tively, Month 6 is prior to dose 3. Error bars = 95% CIs. DISCUSSION the pre-defined non-inferiority criterion Two lots of 2PE-containing hepatitis was not met (group difference = -6.3% B vaccine differing only in the time on the (95%CI -13.6, 0.3)). However, non-inferi- shelf, gave similar immunogenicity when ority between the 2PE New and 2PE Old used for primary vaccination of healthy groups in terms of anti-HBs seroprotection adolescents against hepatitis B. Although rates was established (group difference = non-inferiority of the 2PE New group to 2.2% (95% CI -6.0, 10.4)). Exploratory the HBV(Thio) group in terms of anti-HBs

880 Vol 41 No. 4 July 2010 STABLE IMMUNOGENICITY OF HBV VACCINE

30 2PE New ther removal of thiomersal from 2PE Old the vaccines, or the shelf-life of HBV(Thio) Grade 3 the 2PE-containing vaccines ap-

20 peared to influence the inci- dence of adverse events after vaccination. Grade 3 symptoms after vaccination occurred infre- 10 Percentage of doses quently and were reported after no more than 0.2% of doses. No SAEs reported during the entire 0 study period were considered Pain Redness Swelling Fatigue Gastrointestinal Headache Fever by the investigator to be related Fig 3–Solicited local and general solicited symptoms accord- to vaccination. These results are ing to the protocol safety cohort. N = 415 doses in the in line with a study conducted 2PE New group, N = 408 doses in the 2PE Old group with another preservative-con- and N = 403 doses in the HBV(Thio) group. taining hepatitis B vaccine seroprotection rate could not be demon- (Rebedea et al, 2006). strated; therefore, the HBV(Thio) group The immunogenicity of 2PE-contain- had a significantly higher seroprotection ing hepatitis B vaccine is stable when stored rate than the 2PE groups. The two groups for <12 months or >18 months, in line with also differed in the number of non-re- data from the WHO (WHO, 2006). The sponders [12 in the 2PE New group and 5 multi-dose hepatitis B vaccine used in this in the HBV(Thio) Group] to the hepatitis study gave acceptable immunogenicity and B vaccine. Another study found the reactogenicity. Complying with the recom- thiomersal free formulations of hepatitis mendations of regulatory authorities to B vaccine had similar seroprotection rates phase out the use of preservatives from when compared to thiomersal containing vaccines (AAP, 1999; EMEA, 1999), hepatitis B vaccine (Van Damme et al, GlaxoSmithKline Biologicals has discontin- 2009). ued production of multi-dose versions of An exploratory analysis showed no Engerix™-B vaccine and has removed all difference between the 2PE-containing preservative from monodose Engerix™-B groups in terms of anti-HBs seroprotection formulations. rates or GMCs one month after dose 3.The kinetics of the responses over the 3-dose ACKNOWLEDGEMENTS vaccination course were within the same range for both groups (Fig 2). Data from The authors thank Dr Joanne Wolter the WHO shows hepatitis B vaccines are and Geetha Subramanyam for medical generally stable for up to four years when writing and Dr Roselynn Tien for edito- stored at 2-8ºC (WHO, 2006). Results from rial assistance and manuscript coordina- the present study also suggest the shelf- tion. Joanne Wolter is a freelance medical life of the 2PE preserved hepatitis B vac- writer; Geetha Subramanyam and Roselynn cine did not have an impact on vaccine Tien are employees of GlaxoSmithKline immunogenicity for >18 months. Biologicals. The vaccines had a clinically accept- GlaxoSmithKline Biologicals was the able safety and reactogenicity profile. Nei- funding source and was involved in all

Vol 41 No. 4 July 2010 881 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH stages of the study and analysis. Glaxo able from: URL: http://www.ema.europa. SmithKline Biologicals also funded all costs eu/pdfs/human/bwp/251700en.pdf associated with the development and pub- Keating GM, Noble S. Recombinant hepatitis B lishing of the present manuscript. The cor- vaccine (Engerix-B): a review of its immu- responding author had full access to the nogenicity and protective efficacy against data and was responsible for submission hepatitis B. Drugs 2003; 63: 1021-51. of the publication. Poovorawan Y, Sanpavat S, Pongpunlert W, et al. Comparison of a recombinant DNA Bernard Hoet, Hans L Bock and hepatitis B vaccine alone or in combina-

Karthik Srinivasa were employees of tion with hepatitis B immune globulin for GlaxoSmithKline Biologicals at the time the the prevention of perinatal acquisition of study was conducted and during manu- hepatitis B carriage. Vaccine 1990; (suppl script preparation. Nina G Gloriani de- 8): S56-9. clared she has no conflicts of interest. The Poovorawan Y, Sanpavat S, Pongpunglert W, laboratory tests were all performed by et al. Long term efficacy of hepatitis B vac- GlaxoSmithKline Biologicals. cine in infants born to hepatitis B e anti- Engerix-B is a trademark of the gen-positive mothers. Pediatr Infect Dis J 1992; 11: 816-21. GlaxoSmithKline group of companies. Rebedea I, Diaconescu IG, Bach D, Bartelsen O, Arndtz N. Comparison of thiomersal- REFERENCES free and thiomersal-containing formula- American Academy of Pediatrics. Joint State- tions of a recombinant hepatitis B vaccine ment of the American Academy of Pedi- (Hepavax-Gene) in healthy adults. Vaccine atrics (AAP) and the United States Public 2006; 24: 5320-6. Health Service (PHS). Pediatrics 1999; 104: Van Damme P, von Sonnenburg F, Hatz C, Hoet 568-9. B, Lefevre I, Leyssen M. Long-term im- Centers for Disease Control and Prevention munogenicity of preservative-free hepa- (CDC). Update: expanded availability of titis B vaccine formulations in adults. J thimerosal preservative-free hepatitis B Med Virol 2009; 81: 1710-5. vaccine. MMWR Morb Mortal Wkly Rep World Health Organization (WHO). Safe vac- 2000; 49: 642-51. cine handling, cold chain and immuniza- Clements CJ, Wesselingh SL. Vaccine presen- tions: A manual for the Newly Indepen- tations and delivery technologies-what dent States. Geneva: WHO, 1998: 1-23. does the future hold? Expert Rev Vaccines World Health Organization (WHO). Hepatitis 2005; 4: 281-7. B. Fact sheet no.204. Revised October EMEA. European Agency for the Evaluation 2000. (Cited 2010 Mar 2). Available from: of Medicinal products (EMEA). 1999. URL: http://www.who.int/mediacentre/ EMEA public statement on thiomersal factsheets/fs204/en/index.html containing medicinal products. EMEA World Health Organization (WHO). Hepatitis 2001; publication No. 20962/99. (Cited B vaccines. WHO position paper. Wkly 2010 Mar 2). Available from: URL: http:// Epidemiol Rec 2004; 79: 255-63. www.ema. europa.eu/pdfs/human/press/ World Health Organization (WHO). Tempera- pus/2096299EN.pdf ture sensitivity of vaccines. Geneva: EMEA Committee for Proprietary Medicinal WHO, 2006: 1-62. products. Points to consider on the reduc- Zanetti AR, Van Damme P, Shouval D. The glo- tion, elimination or substitution of bal impact of vaccination against hepati- thiomersal in vaccines: CPMP/BWP/2517/ tis B: a historical overview. Vaccine 2008; 00: 2001: 1-5. (Cited 2010 Mar 2). Avail- 26: 6266-73.

882 Vol 41 No. 4 July 2010